index

1. Workflow

• You can use the scroll wheel to zoom in on the graph to view a specific element.

Code

library(dplyr)
targets::tar_visnetwork()

Le chargement a nécessité le package : geometry
Le chargement a nécessité le package : ape
Le chargement a nécessité le package : rcdd
If you want correct answers, use rational arithmetic.
See the Warnings sections in help pages for
    functions that do computational geometry.

2. Data

• summary of \(\delta\)¹³C and \(\delta\)¹⁵N data by species

Code

isotope_data <- targets::tar_read(isotope_data_fish) %>%
  as.data.frame() %>%
  select(species, d15n, d13c) %>%
  group_by(species) %>%
  summarise(
    n = n(),
    mean_d15N = round(mean(d15n),2 ),
    sd_d15N = round(sd(d15n),2),
    mean_d13C = round(mean(d13c), 2),
    sd_d13C = round(sd(d13c), 2)
  )

htmltools::tagList(DT::datatable(isotope_data))

3. Isotopic niches

Ellipses

• Standard ellipses at 40% of the 16 fish species with all sampling depths combined
• Δ \(\delta\)¹³C = 2.36‰
• Δ \(\delta\)¹⁵N = 5.94‰

Code

targets::tar_load(species_niche)
knitr::include_graphics(species_niche)

Overlaps

• Asymmetric matrix of isotopic niche overlap between the different species sampled at all sampling depths combined. The isotopic niche was estimated using a 40% ellipse. The overlap values lie between 1 (i.e. the niche of species 1 at the bottom of the matrix is completely covered by the niche of species 2 on the left of the matrix) and 0 (i.e., the niches of the two species are completely separate).

Code

targets::tar_load(plot_matrix)
knitr::include_graphics(plot_matrix)

4. Depth segregation

Clusters

• input data for clusters = overlap matrix

Code

targets::tar_load(nb_cluster_gs)
knitr::include_graphics(nb_cluster_gs)

• It is not ecologically relevant for a species to be alone in its cluster, so 5 clusters are selected.

Code

cluster <- targets::tar_read(cluster_definition) 

htmltools::tagList(DT::datatable(cluster))

Depth distribution

• From complete 2021 trawling data (not only individuals sampled for isotopic analyses)

Code

targets::tar_load(depth_distribution)
knitr::include_graphics(depth_distribution)

5. Null models

Isotopic species niche size

• Comparison of the isotopic niche area of each species in the different depth layers. Density plots represented the distributions of the estimated isotopic niche area based on the resampling of isotopic values (i.e. null model). The actual niche of each species is represented by the dotted vertical line

Code

targets::tar_load(niche_area_sp)
knitr::include_graphics(niche_area_sp)

Sum of isotopic niche overlaps

• Comparison of cumulative isotopic niche overlap standardized by the number of species in each depth layer. The distributions represent the cumulative overlap of isotopic niches based on the resampling of isotopic values (i.e., null model). The dashed lines represent the cumulative isotopic niche overlap observed in each depth layer.

Code

targets::tar_load(overlap_sp)
knitr::include_graphics(overlap_sp)

6. Isotopic diversity index

• IDiv = divergence, IDis = dispersion, IEve = evenness and IUni = uniqueness

Code

diversity_index <- targets::tar_read(diversity_index) %>%
  filter(index %in% c("IDiv", "IDis", "IEve", "IUni")) %>%
  mutate(ID_scl_ab = round(ID_scl_ab, 4)) %>% 
  tidyr::pivot_wider(names_from = index, values_from = ID_scl_ab)

htmltools::tagList(DT::datatable(diversity_index))

PCA

Code

targets::tar_load(PCA)
knitr::include_graphics(PCA)

7. Appendices

Appendix A

• Standard ellipses at 40% of the trophic guilds obtained by clustering based on species \(\delta\)¹³C and \(\delta\)¹⁵N values

Code

targets::tar_load(niche_cluster)
knitr::include_graphics(niche_cluster)

Appendix B

• Graphical representation of all the indices in each depth layer
• A= epipelagic, B= upper-mesopelagic, C= lower-mesopelagic, D= bathypelagic and E= bottom-proximity
• IDiv = divergence, IDis = dispersion, IEve = evenness and IUni = uniqueness

Code

knitr::include_graphics("figures/isotopic_diversity_indices.png")